Our long-term goal is to reveal benchmarks and best practices for end-of-life (EOL) nursing care that support patients'dignity and comfort. Millions of dollars are spent each year on treatments of questionable value to hospitalized patients who are near death. The maturation of the electronic health records (EHRs) and standardized terminologies now make it possible to capture data and use it to determine the benchmarks and best practices of effective care or risks for adverse events. In our preliminary study of an extensive HANDS database of actual nursing care practices with 39,322 care episodes, we identified 1,394 EOL patients and we examined their "pain care" to demonstrate the feasibility of our methods and adequacy of power to determine effective pain benchmarks. This study yielded 4 statistically significant and clinically important benchmarks: (1) 51% of EOL patients failed to meet expected pain outcomes at discharge to hospice or death;(2) pain control achieved in the first 24 hours of care was predictive of pain control for the entire stay;(3) EOL patients with both cardiopulmonary and pain diagnoses had significantly poorer pain outcomes than those who do not;and (4) certain interventions were more likely to achieve pain control. We translated these pain benchmarks into an EHR user interface that is now ready for usability testing by a prospective sample. Building on these important findings, we propose to identify benchmarks and best practices for 5 additional care problems and 5 combinations of problems that are common to hospitalized EOL patients. From the validated dataset of care plan histories for 1,394 EOL patients from 8 different acute care units in 4 Midwestern hospitals, we will characterize EOL care using the following attributes: 1) patient and provider demographics;and 2) the nursing diagnoses, interventions, and outcomes as they evolved during each patient's care episode. Once we determine the benchmarks, then we will build a benchmarking system that presents the findings visually on EHR computer screens that help nurses to see and use the benchmarks as they plan and document care in the EHR. Our specific aims are to: Aim1. Determine patterns among all attributes recorded in HANDS for meaningful associations among the attributes that reveal benchmarks and best practices. Aim 2. In a diverse sample of 75 nurses, determine the usability (accessibility, visual appeal, content utility to guide intended use at the point of care) of new visualization screens that display the benchmarks for outcome pattern ratings (at the level of the person, unit, hospital, and across the 4 hospitals) and best practices. Study findings will inform future prospective practice-based research to verify the effectiveness of the benchmarking system to produce desired outcomes for hospitalized EOL patients. Such findings are urgently needed to enable best practice nursing care that supports hospitalized EOL patients'dignity and comfort as they die. PUBLIC HEALTH RELEVANCE: In this study, information that nurses document in EHRs will be used to discover the best treatments to support the dignity and comfort of dying patients and their families. We will then take the new knowledge and build and test EHR innovations that will make sure nurses use the knowledge when caring for dying patients.